phildenhoff

What is a submarine article?

Unfortunate to see an otherwise-good article padded by AI-isms ("But here’s the twist: that search I described earlier was never sent to botd. All the history was in SQLite, so I pointed another agent at it and got the analysis anyway. The tool died; the data didn’t.", amongst other examples).

Maisem Ali writes about how they have not stopped engineering despite handing production of code over to AI. I think that's an interesting perspective for a few reasons, but most of all, I'm not sure I agree.

First, we see a prime example of how that can blow up. "botd died this month. It crumbled under its own weight." It's hard to imagine how an agent harness (or is it an overlay on top of other harnesses?) could crumble under its own weight from sheer technical complexity. What makes a tool like this "crumble"? If properly architected, it seems that any outside change could be adaptable. Failing tests could be fixed.

This AI coding evolution has given us the ability to spin up prototypes we don't understand the inner workings of, but I believe for anything of value (and, I would argue, botd appeared to have value to Maisem) we should at the very least understand and influence the architecture and engineering of what we're building. If Maisem had done that for botd, it would not and _could not_ have "crumbled under its own weight". Whatever outside influence required change within botd would have been manageable.

Second, the off-loading of the production of this article indicates to me a lack of respect (with no offence to Maisem, because this is what much of the industry is doing today) for the value in thinking through the writing process. There's a lot that I've learned personally about software I worked on that only came to me through trying to write explainers. Taking a step back from the software we build and looking at it from a birds-eye view, to share how it is constructed or how to use it, helps develop our understanding.

Maisem's article is already very good, and I hope that is demonstrative of their own writing! I would read more! But I want to read words written by a person for other people. AI can help edit, it can help us build ideas, but in code and in our written word, AI should not be _the author_.

Of course, now you’ve piqued my interest - what is it that you want from your library management tool?

FWIW the tool I’ve built has no built-in ereader because I use apps like Apple Books or my kobo to read. i don’t need another half-built ereader

I love it! Thanks for sharing. That makes perfect sense - available from anywhere with limited infrastructure.

Looking forward to seeing what comes next for Bookshelf!

Having built a tool in this space, I enjoy seeing all the Calibre alternatives that have been coming out in the last few years! It's fascinating to see the specific wants that people have.

Bookshelf, for example -- what possesses someone to make an ebook library that _only works on object storage_? Is this a weekend project, answering the question of "what would this look like"? Or was it something the author wanted for a long time?

It's not _theft_ because nothing is _stolen_. The tokens are being used in a way that breaches the contract agreed to by whomever set up the account with OpenAI, Anthropic, Kilo, Antigravity etc. but it's not theft.

Fun idea and a great easter egg, but as a game, the tutorial needs some work. I was so lost for so long and I'm still not sure I understand half of the game!

You’re expecting the wrong thing. The demo demonstrates the insane inference rate of dedicated hardware. Iirc it’s llama 3 or something. Not a very good model by today’s standards. But it runs at 16k tokens per second, an order of magnitude above the competition.

Imagine what’s possible if you had GLM-5.2 turned into a hardware chip like this.

Very cool -- I often thought that Unisons caching approach should be adopted more widely. Look forward to poking at it later!

The company I work for, Digits, has been regularly updating our AI-vs-human bookkeeper benchmark. Look at page 8 -- many models are nearly as accurate as a human bookkeeper

https://digits.com/downloads/beyond-the-hype-evaluating-llms...

Rocket lab used to be a New Zealand source of pride, having started there. From the press release, now it’s American. What happened?

What would have happened, in that first story, if he had left the ship and swam to a passing boat? Or swam to shore? He was apparently able to leave as later in his imprisonment, the boat drift closer to shore and he swam from there. Why not just leave?

I don't think this car is designed for them, similar to how trucks are not designed for urban areas. Not every car has to cater to every demographic.

If you have a charger at home, or at work, why does it matter if you have to charge it daily or weekly? Yes, for some lifestyles, range matters. For others, a 125 km range is perfectly acceptable

Hey -- we've loved using Conductor here at Digits. I've been using it since February and only recently swapped to using Orca for Remote SSH and perf improvements. Looking forward to your general release of Conductor Cloud!

I've been wondering, since you're building desktop software, how do you get AI to test your changes? Boot the whole app? Run the frontend/UI with a mock backend?

How often do you use it?

Immediately I started thinking how nice it would be to use natural language to have LLMs generate a deterministic filter for stories matching content I DO care about, filtered from New. Instead of filtering it out.

I think this is an excellent consideration of how the principles of Zen and the Art of Motorcycle Maintenance can apply to programming. I've always strived for Quality in my work, and I struggle today to balance that Quality with speed as I leverage coding agents more and more. I'm able to build whatever I want -- the gap is my domain expertise and taste. Before, my domain was "building and maintaining excellent software" but, these days, that's not enough.

Do you manually translate or translate with an LLM? While reading, I was wondering how common these kinds of written tics are in languages outside English.

Nice project! Curious why build a new front end instead of fixing (?) the behaviour in upstream Kakoune? I think the original author’s still working on Kakoune

Having built some stuff with Tauri, being able to debug using Chrome instead of a Safari/Webkit console would be _so nice_.

But it is a browser -- just not a web browser

I would not recommend driving over the Juan de Fuca strait

> 90,000 people live in the core, which is about the maximum daytime population of Disneyland. More people live in the greater capital district (400,000), but because Victoria is subject to very particular microclimates, the further you get from the city itself, the worse the weather becomes

True, but outside of the City of Victoria is Saanich, the largest municipality of the 13 that make up the Capital Regional District. Saanich has roughly the same climate and another ~110k.

Also, anecdotally, despite the smaller population, Victoria is a tourist destination and swallows cruise ships worth of passengers with ease. If you want to visit, come visit. You'll find all the amenities you need.

They created a fascinating tool -- I'll be following your progress to see what you build!

I agree that it shouldn't mysteriously be a pointer all the time, but using a pointer for the cursor is a preference, not a standard. In native apps (outside the web), the cursor is often not turned into a pointer when hovering clickable content.

I hope qntm has the chance to traditionally publish Ra and have it edited as well. I enjoyed the book a lot, but felt it needed a solid once over.

Really enjoyed the novel though! Planning to reread it in the spring.

For those coming to this thread after it was scrubbed, the author unfortunately felt they had to shut the project down after being pressured by their employer. The name is redacted and the website shut down.

For what it's worth, commentary from others on the JJ Discord suggested that this could not be a legally binding requirement as the author is located in California. California has laws that prevent employers from controlling employees inventions outside of work hours when using their own devices: https://codes.findlaw.com/ca/labor-code/lab-sect-2870/

Why not set it up as your default search engine? That’s what I did. Hard to forget when it’s the default

Are you upset about DRM in general? Or that Signal, by default, prevents Windows from capturing the Signal window when it screenshots the screen every few seconds?

because it sounds like Windows is the problem here, doing this screenshotting at all. And Signal allows you to disable the anti-screenshotting measure

andes314, can you expand on how you see this as Postman for MCP?